Remote Sensing Applications 113

Figure 2.92  Multispectral classification of Augsburg in Germany. (Land cover classi-

fication, courtesy of GAF Remote Sensing and Information Systems, Munich, Germany.)

Provided that the images are geometrically registered with respect to each
other, multitemporal and multisensor data fusion may also be used. A compar-
ison of multitemporal data is generally done for change detection. Here again,
a knowledge-based approach is advantageous.

REMOTE SENSING APPLICATIONS

A great number of satellite systems have provided satellite imagery for remote
sensing applications in different disciplines. Table 2.11 provides a summary of
the classical remote satellite systems with optical sensors. Table 2.12 augments
this list with the high repletion satellites, which have been available since 1999.
Table  2.13 lists important new medium resolution satellites that have been
launched since the millennium. Table 2.14 shows the early classical radar satel-
lite systems; Table 2.15 augments the list with the recent high-resolution radar
systems since the millennium.
In addition to these satellites, a great number of images have been acquired
through shorter duration missions by Russian photographic camera systems from
Kosmos satellites and the MIR space station, for example, the Russian KFA 1000,
KVR 1000, and TK 350 cameras; the German Metric camera on the space shuttle;
as well as the MOMS- 2P digital sensor flown on the space shuttle and on MIR.
 114
Remote Sensing
TABLE 2.11  COMMONLY USED CLASSICAL OPTICAL REMOTE SENSING SATELLITES
Repeat
Type Name Orbit Cycle Swath h Resolution Bands Country Application
Meteorological Meteosat Geostationary 30 min Half spheric 36000 5 km ESA Meteorology,
km climate
GOES-1 75°W USA
GOES-2 135°W USA
GSM 140°E Japan
Insat India
Meteor 3 Russia
FY 2 105°E China
Meteorological NOAA Polar 12 h 2394 km 705 km 1 km 5 USA Climate
DMSP Polar 2400–3000 3 km 0.58–21.5 USA Military
km μm
5 satellites Polar 6 Russia Climate
Meteor P 6 (TOMS, ozone)
0.31–0.38
μm
Earth resources Landsat Sun synchronous 18 d 185 km 918 km 80 m 3–4 USA 1972–1984
(1–3) MSS 9:30 at 40°
 Landsat 10:30 at 40° 16 d 185 km 705 km 30 m (thermal 7 USA 1982, 1984
(4–5) TM 120 m)
Landsat 7 10.00 at 40° 16 d 185 km 705 km 15 m pan 7 USA 1999
TM
30 m MS
Earth resources Spot P 1–4 Sun synchronous 26 d 60 km 832 km 10 m pan 1 France 1986
Spot XS 1–4 10:00 at 40° 20 m MS 3 1993
Earth resources JERS 1 OPS 75 km 568 km 20 m 7 Japan 1992
Earth resources IRS 1 A, B Sun synchronous 22 d to 141 km 904 km MS 36.6 m 3 India 1988, 1991
24 d
IRS 1 C, D 9:25 at equator 774 km Pan 5.6 m 1 1995, 1997
MS 23.5 m 3
WIFS 188 m 2
Cartographic Ikonos 2 Sun synchronous 11 km 677 km 1 m pan 1 USA 1999
EROS A1 Sun synchronous 12 km 480 km 4 m MS 1 Israel 2000
Sun synchronous 8 km 450 km 1.8 m pan 1 USA 2001
0.6 m pan

Remote Sensing Applications

2.4 m MS

115
 116
Remote Sensing
TABLE 2.12  RECENT OPTICAL HIGH-RESOLUTION SATELLITE SYSTEMS
Number
Agency Satellite Year GSD-pan GSD-MS of Bands Swath Remarks
Digital Globe, USA Ikonos Sept 1999 0.82 m 3.2 m 4 11.3 km
Imagesat, Israel EROS-A Dec 2000 1.9 m — — 14 km Pan only
Digital Globe, USA Quickbird Oct 2001 0.61m 2.44 m 4 16.5 km
CNES, France Spot 5 May 2002 2.5/5 m 5/10 m 4 60 km
Taiwan Formosat 2 May 2004 2m 8m 4 24 km
ISRO, India Cartosat 1 May 2005 2.5 m — — 30 km Pan, stereo
Jaxa, Japan Alos Jan 2006 2.5 m 10 m 4 70 km Stereo
Imagesat, Israel EROS-B April 2006 0.7m — — 11 km Pan only
Russia Resurs DK1 June 2006 0.9 m 2.5 m 3 28.3km
Digital Globe, USA WorldView 1 Sept 2007 0.5 m — — 17.6 x 14 km pan
(750000 km2/d)
ISRO, India Cartosat 2A,B April 2008 (July 0.8 m — — 9.6 km Pan only
2010)
Digital Globe, USA Geo Eye 1 Sept 2008 0.41 m 1.65 m 4 12.5 km (250000
km2/d)
Digital Globe, USA WorldView 2 Oct 2009 0.46 m 1.8 m 8 16.4 km (1 M
km2/d)
Astrium, F & D Pleiades1A(B) Dec 2011 (Dec 2012) 0.7 m (0.5 m 2m 4 20 km
resampled) constellation
China ZY-3 Jan 2012 2.1 m 3.5 m 2 51 km Stereo
Korea Kompsat 3 May 2012 0.7 m 2.8 m 4 16.8 km
TABLE 2.13  RECENT OPTICAL MEDIUM-RESOLUTION SATELLITE SYSTEMS
Number
Agency Satellite Year GSD-pan GSD-MS of Bands Swath Remarks
ISRO, India IRS-1C Dec 1995 5.8 m 23 m 4 141 km
NASA, USA Landsat 7+ETM April 1999 15 m 30 (90) m 7 185 km
NASA, USA ASTER Dec 1999 — 15m VNIR 1–3 60 km
— 30m SWIR 4–9
— 90m TIR 10–14
China-Brazil CBERS 2 Oct 2003 20 m to 260 m 4 113 km
Germany RapidEye Aug 2008 6.5 m 5 77 km Constellation,
red edge, 5
satellites
UK DMC 2 July 2009 2 2m 3 660km
USGS, USA Landsat 8 May 2013 15 m 30 m (100 m 7 185 km
Thermal)
Remote Sensing Applications
117
 118
Remote Sensing

TABLE 2.14  HISTORICAL RADAR SATELLITES

Inclination Country/
Name Year of Orbit Swath h Resolution Polarization Agency
Seasat 1978 72° 100 km 790 km 40 m HH 23.5 cm USA
SIR-A 1981 50° 50 km 250 km 38 m HH 23.5 cm USA
SIR-B 1984 58° 40 km 225 km 25 m HH 23.5 cm USA
SIR-C 1994 51° 30 to 60 km 225 km Variable 13–26 m Multiple HH, HV, USA
VH, VV, 23.5
cm, 58 cm, 3.1
cm
ERS 1/2 1991, 1995 Polar 100 km 785 km 30 m VV 5.7 cm ESA
JERS 1 1992 Polar 75 km 568 km 18 m HH 23.5 cm Japan
Almaz 1991 Polar 50 to 100 km 350 km 15 m, variable HH Russia
Radarsat Since 1995 Polar 50 to 500 km 800 km Up to 10 m HH 5.7 cm Canada
Envisat (ASAR) 2002 Polar 100 km 800 km 12.5 m HH, VV ESA
TABLE 2.15  RECENT HIGH-RESOLUTION RADAR SATELLITE IMAGING SYSTEMS
Agency Satellite Year GSD Swath Band
ESA ERS-1 1991 10–30 m 100 km C
Jaxa, Japan JERS1 1992 18 m 75 km C
ESA ERS-2 1995 10–30 m 100 km C
Canada Radarsat 1 1995 9–100 m 50–500 km C
NASA, USA SRTM 2000 30 m 225 km C
DLR, Germany SRTM 2000 30 m 45 km X
ESA Envisat 2002 30–1000 m 100–405 km C
Italy, France COSMO-Skymed 2006 1–50 m 10–200 km X
Canada Radarsat 2 2006 3–50 m 20–500 km C
Germany, Astrium TerraSAR-X 2007 1, 3, 18 m 10, 30, 100 km X
China SurveyorSAR 2007 10-25 m 100–250 km C
ISRO, India Risat 1,(2) 2009 (2012) 3–50 m 10–240 km C
Germany, Astrium Tandem X 2010 1, 3, 16 m 10, 20, 100 km X radar interferometry
China Civilian 2012 Radar satellite S
Remote Sensing Applications
119
120 Remote Sensing

Figure 2.93  IRS1C/D image of Munich Airport in Germany. (IRS-PAN/LISS image:

Munich; SI/Antrix/Euromap 1999, GAF 2000, courtesy of GAF Remote Sensing and
Information Systems, Munich, Germany.)

An example image of the airport of Munich, Germany, taken by IRS-C is

shown in Figure 2.93.
These images are available from nationally and internationally operating
space agencies, such as NASA, NOAA, ESA, CNES, NASDA, ISRO, or their vend-
ing agencies (e.g., USGS, Spot Image, Space Imaging, Eurimage, DigitalGlobe).
The data cost still differs greatly but it is in a process of stabilization.
Meteorological data, though reduced in quality, are available over
the Internet free of charge. Original resolution meteorological images can be
obtained at reproduction cost. Medium- and high-resolution images have a
weak to strong commercial component, depending on the policies of the space
agency maintaining the satellite system. Privately funded commercial systems
charge full price, unless they are supported by large government programs.
Global applications, therefore, use low-cost, low-resolution imagery, which
is easily obtainable. Regional and local applications requiring higher resolu-
tions rely on substantial imagery purchases. There, remote sensing competes
with other data acquisition methods with respect to obtainable quality, cost,
and evaluation effort and time.
Project-based research applications were the easiest to be realized. The
present focus of applications is to concentrate on organized data acquisition
 Remote Sensing Applications 121

and analysis programs depending on the socioeconomic priorities to be placed

on applications made possible by public or industrial funding.
We will now consider the situation in the major application areas of
remote sensing.

Meteorology and Climatology

Atmospheric sciences study the different layers of the earth’s atmosphere:
•• Troposphere, from 0 to 20 km altitude
•• Stratosphere, from 20 to 50 km altitude
•• Mesosphere, from 50 to 80 km altitude
•• Thermosphere, from 80 to 300 km altitude
The “weather zone” is the troposphere, which is of direct meteorological inter-
est. However, the other zones also affect weather and climate.

Ozone
The earth’s ozone shield extends from an altitude of 25 to 60 km. It absorbs
or reflects most of the ultraviolet (UV) light, so that only minimal amounts of
ultraviolet reach the earth’s surface.
NASA has launched an ultraviolet sensor, TOMS (Total Ozone Mapping
Spectrometer), which observes the ozone layer in six bands between 312.5 mm
and 380 mm wavelength. From these six bands, three pairs can be formed. They
determine transmission minus absorption of UV energy. This ratio is a mea-
sure of ozone concentration. TOMS was carried on the U.S. satellite Nimbus 7
from 1978 to 1993, on the Russian satellite Meteor 3 from 1991 to 1994, and on a
special satellite since 1996. TOMS detected a rapidly deteriorating ozone con-
centration over the south polar regions in 1992 and 1993, which created great
public interest. The cause could have been the eruption of the volcano Mount
Pinatubo in 1991, even though aerosols produced by human activity may also
have played a part. Since then, ozone measurement has been a major global
remote sensing application.
The European Space Agency (ESA) launched an ozone sensor on the ERS-2
satellite in 1995. Figure 2.94 shows the global ozone concentrations on a par-
ticular day (Figure 2.94a) and a monthly average (Figure 2.94b). The observa-
tions were continued by Schiamachy on Envisat under the GOME experiment
for ozone studies. Figure 2.95 shows the ozone vertical column density for the
northern hemisphere and Figure 2.96 for the southern hemisphere.

Cloud Mapping
The first U.S. imaging satellite launched in April 1961 was Tiros 1, which
made it possible to observe clouds. Today, geostationary satellites permit a
122 Remote Sensing

(a) (b)

Figure 2.94  Images of ERS2 ozone sensor of the globe. (ERS-2/GOME; DLR, courtesy
of DLR, Oberpfaffenhofen, Germany.)

daily weather watch following the movement of cloud patterns. The satellites
Meteosat (over Africa), GOES 1 (over Venezuela), GOES 2 (over Hawaii), GMS
(over the Philippines), and Insat (over the Indian Ocean) gather images every
30 minutes in the visible and thermal range. These may be geocoded and used
in animations, which are commonly shown on television programs.

Figure 2.95  Image of Envisat Schiamachy GOME ozone sensor of the Northern
Hemisphere. (From DLR, Oberpfaffenhofen, Germany.)
 Remote Sensing Applications 123

Figure 2.96  Image of Envisat Schiamachy GOME ozone sensor of the Southern
Hemisphere. (From DLR, Oberpfaffenhofen, Germany.)

The data collected at an interval of 6 hours may be used to determine param-

eters of the radiative transfer model, into which the distribution of clouds,
water, ice, snow, and the land mass is entered.
The combination of visual and thermal bands permits the visual separa-
tion of clouds, water, ice, snow, and land. If combined with atmospheric non-
remote sensing measurements, a radiative transfer model can be arrived at.
Figure 2.97 shows the thermal GOES-1 image of a hurricane, and Figure 2.98
gives a view of a hurricane with SeaWiFS from OrbView-2.

Rainfall
The measurement of rainfall is of great meteorological interest. The sources of
worldwide rainfall data are rain gauges, which are very scarcely distributed
over the globe. If no rain gauge data are available over a region, data from
Meteosat, GOES, and so on relating to thermal bands can be used to determine
cloud temperature. Cold clouds with <235 K temperature give an indication of
possible rainfall.
On the U.S. military DMSP satellite, four wavelengths from 0.35 to 1.55 cm
are provided for passive microwave sensing in two polarizations (HH, VV). The
images of 55 km ground resolution permit the derivation of a scattering index,
indicative of rainfall.
124 Remote Sensing

Figure 2.97  Thermal GOES image of hurricane. (Meteosat-3 MVISSR; Eumetsat, pro-
cessed by DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)

Figure 2.98  SeaWiFS OrbView-2 image of a hurricane. (From SEOS. EARSeL is a part-
ner in the ESA project SEOS, Greece.)
 Remote Sensing Applications 125

Wind
Since wind drives ocean currents, scatterometers can measure the roughness
of the sea to estimate wind vectors. Radar images are also able to detect rough-
ness parameters.

Weather Prediction
Terrestrial measurements for weather forecasting can easily be combined with
remote sensing data for cloud motion, the estimated precipitation, and the
measurement of surface temperature.
Other phenomena detected from images are
•• Analysis of snow cover
•• Location and the motion of tropical storms
•• Detection of fog and its dissipation

Climate Studies
Climate studies become possible by the comparison of NOAA-AVHRR aggre-
gates on a seasonal and annual basis.

Oceanography
The geodetic aspects to be studied are ocean heights, as determined in the ESA-
ERS1 Topex mission by radar altimeters. When related to the geoid, the data
permit the derivation of ocean height (see Figure 2.99).
Let us now consider some phenomena of direct interest to remote sensing.

Figure 2.99  Heights from Topex altimeter ocean ERS-1. (Courtesy of DLR,
Oberpfaffenhofen, Germany.)
126 Remote Sensing

Spectral Radiance µWch–2 sr–1 mm–1

60 mg/m3
1

8 mg/m3
0,1

0,01
1 mg/m3

0,001
0,4 0,7
Wavelength in µm

Figure 2.100  Spectral chlorophyll response at sea.

Ocean Productivity
The main objective of studying ocean productivity is to detect organic sub-
stances, such as phytoplankton, which are important for fisheries. It contains
chlorophyll, which can be differentiated from suspended sediments prevalent
near the coast and transported by estuaries (see Figure 2.100). Figure 2.101 shows
an image of chlorophyll concentration and sediments in the Strait of Gibraltar.

Figure 2.101  Chlorophyll separation of pigments, sediments, and aerosols from MOS.
Images of IRS for the Strait of Gibraltar. (Left) Multispectral image followed by the sepa-
ration of pigments (chlorophyll concentrations), sediments, and aerosol concentrations.
(Right) A black-and-white image for the separation of clouds over the sea. Pigments in
µg/l (blue = 0, green = 3, red = 6; sediment: blue = 0, red = 5; aerosol optical thickness:
blue = 0, red = 1). (MOS-IRS; DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)
 Remote Sensing Applications 127

Figure 2.102  Algae bloom in the Bay of Bisquay. (From SEOS. EARSeL is a partner in
the ESA project SEOS, Greece.)

Sediments reflect mainly in red. Therefore, a blue/green ratio can indicate

the chlorophyll concentration at sea (see Figure 2.102 for algae bloom). The
actual concentration can be calibrated by in situ measurements.
The satellites Nimbus 7 and OrbView-2 carried the sensor SeaWiFS for
observation in eight channels at 1 km resolution with a swath of 2800 km.
Of particular interest is the observation of sea surface temperature, which
is made available at weekly intervals (see Figure  2.103 for the eastern
Mediterranean Sea).

Ocean Currents
Ocean currents are visible along the coast because of plumes of suspended
matter. In the midocean, the radiant temperature, which can be measured day
and night by NOAA-AVHRR, shows the distribution of ocean currents. Global
thermal phenomena, like El Niño, can be monitored by NOAA satellites (see
Figure 2.104).
Radar images are also an indicator that can help to determine the level of
surface roughness, since currents produce small waves.

Sea Ice
The principal objective of the Canadian Radarsat satellite is to be able to
follow sea ice motion in the polar areas. In multitemporal mode the images
can be observed in stereo, giving an indication of the direction of flow (see
Figure 2.105).
128 Remote Sensing

Figure 2.103  Sea surface temperature in the Eastern Mediterranean, weekly mean
temperature, summer 1994. Dark red = 30°C, blue = 23°C. (From NOAA-AVHRR; DLR,
courtesy of DLR, Oberpfaffenhofen, Germany.)

The ESA-ERS1/2 satellites have been able to classify Arctic sea ice. The
thermal band six of Landsat TM can distinguish ice temperatures. Thin ice is
warmer than thick ice.
The surface roughness of sea ice can be measured by nonimaging radar
scatterometers, flown from aircraft at altitudes below 1 km.

Figure 2.104  El Niño observed sea surface temperature anomaly by ERS along track
scanning radiometer (monthly mean, October 1997). (From CEOS CDROM 98, courtesy
of CNES, Toulouse, France.)
 Remote Sensing Applications 129

Figure 2.105  Radarsat image of icebergs on the west coast of Antarctica with wave
patterns around the icebergs. (From ERS-1, © ESA, processed by DLR, courtesy of DLR,
Oberpfaffenhofen, Germany.)

Bathymetry
Bathymetry at sea is generally made by sounding from ships, which have, how-
ever, difficulties in navigating in shallow areas. In these areas, remote sens-
ing from aircraft and satellites can help to assess water depth, even though,
depending on the turbidity of the water, light penetration and reflection is
generally limited to not more than 10 m water depth. Figure 2.106 illustrates
the transmission in water, which is best for the green band. A blue–green
ratio can therefore be used for an assessment of water depth in shallow areas.
Figure 2.107 shows the composition of the Wadden Sea near Wilhalmeshaven,
Germany, at low tide.

Environment
Remote sensing concentrates on the detection of environmental pollution.

Hazardous Waste
The task to identify hazardous waste is to find its location, to map it, and to
monitor it. If identification is not directly possible by visual inspection, the
health of vegetation over the waste area can be a good indicator for hidden
130 Remote Sensing

100

Ocean
Transmission in %
10
Coast

1 Estuaries

0,1
0,4 0,7
Wavelength in µm

Figure 2.106  Water penetration in the visual spectrum.

Figure 2.107  Wadden Sea on the German North Sea coast near Wilhelmshaven at low
tide. (IRS 1D-LISS image, SI/Antrix/Euromap 1998, GAF 2000, courtesy of GAF Remote
Sensing and Information Systems, Munich, Germany.)
 Remote Sensing Applications 131

waste. A ratio of infrared–red is a good discriminator. Waste areas will usually

have a low infrared–red ratio.

Plumes
Thermal pollution in water can be monitored by thermal infrared scanners
operated from aircraft or by multispectral optical sensors from satellites.
Figure  2.108 shows the thermal pollution in the Po River, and Figure  2.109
shows the sediment concentration caused by the Po River in the Adriatic Sea.

Oil Spills
Of particular environmental interest are oil spills at sea and in coastal regions.
Due to its fluorescent properties, oil floating on water has a high reflectance in

Figure 2.108  Po River thermal pollution. (Daedalus, DLR, courtesy of DLR,

Oberpfaffenhofen, Germany.)
132 Remote Sensing

Figure 2.109  Sediment concentration at sea caused by the Po River in the Adriatic
sea. (From MOS-IR; DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)

the ultraviolet region (0.3 to 0.4 μm). An aircraft scanner operated from low
altitude can be used to determine the thickness of the oil film on water.
On a more regional and global level, radar images permit the location of
oil slicks. They dampen the natural roughness of ocean water. They are thus
observable due to their low backscatter in the images. On a local level, laser
scanners are able to differentiate between oil types (see Figure  2.110 as an
example for a ship accident near the coast of Northern Spain). Thermal images
may also be of help, since the oil surface is cooler than the water.

Nonrenewable Resources
Locating nonrenewable resources is the task of geological and geophysical
exploration.
 Remote Sensing Applications 133

Figure 2.110  Oil pollution after ship accident on the north Spanish coast. (From
SEOS. EARSeL is a partner in the ESA project SEOS, Greece.)

Mineral Exploration
For the prospecting of minerals, the methodologies of structural geology are
very important. Geological prospecting is difficult in areas covered by vegeta-
tion, however, it becomes easier in the dry belts of the globe.
Ore deposits are usually associated with structure zones. The interpreta-
tion of lineaments in the rocks permits the visible determination of fractures.
Digital directional filtering of the images helps to better recognize these frac-
tures. Fractures are often also combined with hydrothermal activity. Therefore,
thermal aircraft scanners are useful in this respect. To identify the mineral
composition of outcrops, hyperspectral scanners, such as AVIRIS, offer special
possibilities to distinguish minerals of outcrops.

Oil Exploration
Remote sensing plays an initial role in oil exploration. The analysis of Landsat
and radar images to determine the extent of sedimentary basins is usually the
first step in oil exploration, even though a whole slate of geophysical prospecting
methods must be applied to focus on potential exploration locations. These are
•• Aeromagnetic surveys
•• Gravity surveys
•• Reflection seismology
•• Drilling
134 Remote Sensing

Figure 2.111  Deforestation in Bolivia. (Courtesy of SEOS, Greece.)

Renewable Resources
The task of making an inventory and then monitoring renewable resources is
intimately connected with the establishment of geographic information sys-
tems. It consists of compiling base data by standard mapping procedures, of
supplementing these data by geocoded remote sensing data, and of adding
attribute data from a multitude of sources.
Remote sensing has the advantage of quick acquisition of up-to-date images
at adapted spatial and spectral resolution without limitations of costly ground
access. It has the disadvantage that not all desired categories of information
can be extracted from the images. Figure  2.111 illustrates the deforestation
observed from Landsat in Bolivia.

Land Cover and Land Use

Land cover describes the physical appearance of the earth’s surface, whereas
land use is a land right-related category of economically using the land. Remote
sensing concentrates on observing land cover.
Land cover consists of classifiable terrain objects for which different gov-
ernmental base data providers have implemented a number of object-ori-
ented classification schemes. The UN Food and Agricultural Organization has
devised a hierarchical land cover catalog suitable for the application of remote
sensing and applicable to developing continents, such as Africa. It has been
implemented in the Africover project distinguishing 90 different land cover
classes in different East African countries.
 Remote Sensing Applications 135

The German Surveys and Mapping administration has devised an object-

oriented land cover catalog with the following major categories:

1000 Text information

2000 Settlements
3000 Transportation
4000 Vegetation
5000 Hydrology and water
6000 Topographic elevation data
7000 Administrative boundaries

These major categories are subdivided into subclasses. Many of the subclasses
can be identified and monitored with 85% accuracy by remote sensing at the
global and regional or even at the local level. Figure 2.112 shows the result of
a land cover classification from an airborne multispectral scanner. As shown
in Figure 2.113, the result is significantly enhanced if it is merged with vector
information from a GIS. Figure 2.114 is an example of two satellite image-based
classifications for the island of Tenerife, Canary Islands, with changes of land
cover.

Vegetation
Of particular significance is the monitoring of vegetation. On a global level,
this is done from NOAA-AVHRR images obtainable twice a day. Many of these
images contain clouds, which have to be eliminated from the images. This is
possible by taking all of the images of a 10- or 14-day period and substituting
cloudless pixels into the data set.
Whereas the original 1 km resolution provides too much data, a reduced
resolution of 4 km or 16 km can provide a global vegetation index data set.
This is possible by the use of the ratio for the Normalized Difference Vegetation
Index (NDVI):

Infrared − Red
NDVI =
Infrared + Red

The calculated NDVI for each pixel can attain a value between 1 and –1.
Green vegetation has an NDVI of about +0.7 (shown in red in Figure  2.115),
whereas water, barren lands, and clouds have an NDVI of about –0.3, shown in blue.
After radiometric calibration and the consideration of atmospheric scattering
and geocoding are applied, the NDVI gives a clear indication of seasonal vegeta-
tion changes. On the Northern Hemisphere, greenness tends to rise in May, with
a peak in July and a decrease until September. Yearly comparisons for the respec-
tive months can be useful for comparing crop estimates from year to year.
136 Remote Sensing

Figure 2.112  Daedalus aircraft scanner image over agricultural area. (Daedalus,
© DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)
 Remote Sensing Applications 137

Figure 2.113  Parcel-based land cover from DaedalusScanners image merged with
GIS vector data. (Daedalus; DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)
138 Remote Sensing

Figure 2.114  Land cover changes of Tenerife from supervised classification. (From
SEOS. EARSeL is a partner in the ESA project SEOS, Greece.)

Another use of the NDVI is in the monitoring of tropical vegetation and the
depletion of forests, or of agricultural crops (Figure 2.116 shows the application
for a multispectral aircraft scanner in agriculture).

Natural Hazards
Earthquakes
Seismic risks are not directly observable by remote sensing. However, active
faults may be visible in Landsat images, and the plate movements along these
faults can be monitored by radar interferograms.

Landslides
Fresh landslides are observable in radar images.

Land Subsidence
Land subsidence can be interpreted from images by a change in drainage pat-
terns and an observation of vegetation anomalies.

Volcanoes
Volcanic eruptions are associated with clouds of ash, slope changes, and mud-
flows. Changes in topography can be monitored by radar interferometry, but it
is also possible to use thermal images from Landsat TM and from aerial scan-
ners to study changes in heat emission.
 Remote Sensing Applications 139

Figure 2.115  NDVI of the Oder region from MOS on IRS from channels 7 (651 nm) and
11 (869 nm). (MOS-IRS; DLR, courtesy of DLR, Oberpfaffenhofen, Germany.)

Floods
Floods can easily be monitored with radar images. Figures 2.117 and 2.118 show
multitemporal images of a river flood on the Oder River along the German–
Polish boundary.

Forest and Grass Fires

In fire monitoring, three stages are important:
1. Determination of fire hazards—An index with data indicating humid-
ity, wind speed, cloud cover, ground temperature, and green composi-
tion of the land can be formed to judge fire potential.
2. After a fire has broken out, NOAA-AVHRR can monitor the extent
of the fire in the thermal band. A band 1, 2, and 4 combination can
140 Remote Sensing

Figure 2.116  NDVI from DaedalusScanners for an agricultural area. (NDVI image:
SI/Antrix/Euromap 1997, GAF 1999, courtesy of GAF Remote Sensing and Information
Systems, Munich, Germany.)

Figure 2.117  Radarsat flood image of the Oder River. (© Radarsat 1997, GAF 1997,
courtesy of GAF Remote Sensing and Information Systems, Munich, Germany.)
 Remote Sensing Applications 141

Figure 2.118  Radarsat sequential flood image of the Oder River. (© Radarsat 1997,
GAF 1997, courtesy of GAF Remote Sensing and Information Systems, Munich,
Germany.)

distinguish smoke and differentiate between burned and unburned

areas.
3. After a fire, damage assessment can be made with the help of the
images. Figure 2.119 shows the NOAA image of a forest fire along the
Siberian–Chinese border.
Environmental awareness has prompted the European Space Agency to launch
a series of dedicated satellites for the Global Monitoring for Environment and
Security (GMES) program starting with the launch in 2013. This program was
renamed in 2013, as “Copernicus.”
142 Remote Sensing

Figure 2.119  Forest fire at the Siberian–Chinese boundary. (Image from CEOS
CDROM 98, courtesy of CNES, Toulouse, France.)

The Sentinel satellites focus on:

•• Sentinel 1—Synthetic aperture radar applications in C-band to moni-
tor sea ice, the marine environment, and land surface risks
•• Sentinel 2—A multispectral earth observation system for disaster
relief of the SPOT type
•• Sentinel 3—An ocean-oriented satellite to monitor ocean color, sea
surface topography, and sea temperature